Study Of Models For Port Choice Commerce Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Studied the fitness of Stackerlberg model, the equilibrium model and the fuzzy multiple criteria decision-making models for port choice, and compared the fitness of each model then he summarize the advantage and disadvantage of each model.

In the Stackerlberg Model for Port Choice, the author assumed that the international trade container transportation market could be regarded as a Stackerlberg market. That is, three players including port administrators, carriers and domestic shippers, can be considered in the international container transportation market. Port administrators can be regarded as the superior players, because they have complete information about the optimal behavior of both carriers and domestic shippers under a given port management policy. Carriers, on the contrary, can be regarded as superiors and leaders to domestic shippers who are followers in the market, because carriers have complete information about the optimal behavior of shippers under given carriers' services, and that led to a bi-level Stackerlberg problem.

In this model, the carrier aims to maximize his net revenue by using his strategies of routing, vessel type, call port and frequency of call on each route. The second, model of port of choice, that is the Equilibrium model was based on the assumption that the international trade container transportation market could be regarded as an Equilibrium market. That is, both carriers and domestic shippers aim to maximize their revenues when they choose their ports. According to the structure, the objective function of Equilibrium and constraints are the same as the constraints in the Satckerlberg model for port choice.

The last model was the Fuzzy multiple criteria decision-making method (FMCDM) model for port choice. This model in based on the concept of linguistic variables to describe the human judgments or preference in many situations. Terms of linguistic variables could be called "very poor", "poor", "medium poor", "medium", "medium good", "good", "very good", and so on. These linguistic variables can be also expressed in fuzzy numbers. Owing to the fuzziness of the container port selection problem, the importance weight of various criteria and the preference of each container port are considered as linguistic variables in the paper. These linguistic variables can be expressed in trapezoidal fuzzy numbers. Here the authors refer to interviews with 4 shipping companies in Taiwan. The author found out that the results of the comparison of the Stackerlberg model, the Equilibrium model and the fuzzy MCDM model, Show that these three models cannot be used to explain the actual port choices of carriers and shippers well. Thus there is a need to develop a better model for port choice in future researches.


MAGALA, M., & SAMMONS, A. (2008). A New Approach to Port Choice Modelling. Maritime Economics & Logistics, 10, 9-34.

MAGALA and SAMMONS (2008) looked at the port of choice from the angle that shippers no longer choose a port per se, but rather a supply chain a package or bundle of logistics services.; a pathway to markets - in which a port is just an element albeit an important one of the system. He suggests a new and more effective analytical framework within which the modelling of port choice can be conducted and shipper choice decisions well understood.

The authors argue that the availability and suitability of a particular pathway is governed by a number of spatial, temporal and logistical factors including the availability of a shipping line, land transport, accessibility, connectivity and alignment with shipper needs and strategy. It follows then that when choosing a pathway a shipper will consider not only the possible combinations of ports of origin and destination, but also the availability of shipping lines, routes, land transport and a host of other logistics factors. Since all these elements are part of the pathway, it can be argued with some degree of confidence that the choice of a port is a byproduct of a choice of a logistics pathway in which the total logistics cost is a major supply chain consideration.

Furthermore the authors discuss two key theories the first is based on systems theory and the second on the economic and marketing theory of bundling. The theories share a number of similarities but the way they explain how things work is different though not necessarily contradictory.

The systems theory states that the various parts of the system are linked together and can only be understood by a holistic approach. But in the context of port choice, this means that the choice of a port can be fully understood if the overall choice of a supply chain examined.

Under a systems approach, instead of breaking the supply chain or pathway into its component parts, that is, port, shipping lines, land transport, the parts are integrated into larger, unifying supply chain or pathways framework. This way, from the holistic supply chain view, it is claimed that one can therefore more fully understand and model shipper port choice - albeit one still needs to put the intricate parts together in the first place. In the maritime business this is being done currently by 3PLs and chain integrators.


The way shippers go about constructing and evaluating a set of shipment options is of considerable interest in particular when consumer behavior theories based upon single-item choice fail to predict behavior adequately in a situation involving multiple choices.

In such circumstances bundling appears to offer an adequate framework within which multiple category choice can be well understood. Bundling is concerned with the joint offering of two or more non-substitutable items together at a single, combined price (Yadav, 1994). In the case of a shipping industry, for example, a shipper selects a number of logistics services all of which are necessary to enable them to move the goods from one end of the market to the other.

Under bundling it is predicted that the consumer will choose the bundle with maximum 'utility', that is a bundle that will maximise customer satisfaction for a given service requirement. Thus, bundling is an integrating approach for offering shippers a greater choice of cost control, flexibility, competition, reliability, risk management, and a one-stop service that is more cost-effective than the alternative options.

The modelling approach the author suggested is based on discrete choice modelling and on the treatment of choice of port as an element of a supply chain. The approach views the decision to choose a port-oriented supply chain as a multi-step, complex process which is guided by a variety of economic and non-economic issues, and by quasi-rational assessment of economic costs and benefits that are also filtered through behavioral processes of perception and interpretation.

Two different modelling approaches the first uses discrete choice models based upon revealed choice data. Revealed choice data provides information about past choice decisions individuals made on the subject of interest. The second approach uses choice models derived from stated choice experiment data and it is very useful for situations where the subject of interest is the behaviour in the presence of new situations. A key assumption is that decision-makers behave rationally (although admittedly bounded) and will always choose those alternatives that yield maximum utility or satisfaction.


For the purpose of choice modelling, a supply chain and a port can be described as a combination of factors and attributes at different levels. The key factors or variables relevant to the modelling are those known to influence shipper behaviour and subsequent choices of a port and supply chain. Some variables are qualitative such as the quality of a service or reputation of a port; others such as freight rates, transit time, etc are quantitative.


In this paper it was argued that the current approaches to port choice modeling are at best ineffective at worse outdated and therefore a new approach that models port choice within the framework of a port as an element of a supply chain was needed and was more likely to provide a better understanding of the determinants of port choice.

There is no question that a shipper does choose a port but, in the current business environment in which ports compete as part of supply chain, the approach shippers use is based on selecting a port as an item in a logistics package, often assembled and offered to the shippers by the 3PLs or supply chain integrators which are now becoming the key intervening elements in firm to- firm transaction to effect the freight movement from one end to another. Within this view, it is suggested that discrete choice modelling provides the right modelling framework to handle both the system and the port choice. Particular specification of a model will be context dependent but the universal paradigm is that a port is chosen not in isolation but rather as an element of a supply chain system.


Original Article

An analysis of carriers' selection criteria

when choosing container terminals in


Naima Sa e e d

Department of Economics, Informatics and Social Science, Molde University

College, PO Box 2110 6402, Molde, NO-6402, Molde, Norway.

E-mail: [email protected]

2009 Maritime Economics & Logistics Vol. 11, 3, 270-288

Discussion and Conclusions

In this paper, foreign shipping lines' criteria for selecting container terminals at

two ports in Pakistan have been analysed. This analysis is useful for terminal

operators and port authorities to improve the efficiency and productivity of their

container terminals according to their customers' (carriers') requirements as

well as to compete with their rivals.

For instance, the results indicate that when selecting a terminal, these lines

give priority to attributes, such as the service quality offered at each terminal,

the loading/discharging rate and handling charges. Therefore, terminal operators

should focus particularly on these three factors in order to satisfy their

customers. Night navigation, which is not offered at QICT, is also one of the

important factors for carriers making a decision on a terminal, so in order to

attract more customers, QICT's operator should introduce night navigation in

this terminal.

Furthermore, these lines do not have regard to whether a terminal is private

or under the control of the government. Neither do they consider their personal

contacts or the duration of their relationship with terminal authorities while

selecting these terminals.

Other important findings obtained, after estimation of the linear model, are

that for non-geared vessels and for first- generation vessels, the total stay at

berth is less than that of others. This may be because large, modern vessels

prefer a quicker turnaround time and that is why they had selected the terminal

that offers a fast handling rate and a consequent shorter stay at berth. Therefore,

in order to attract the large, modern vessels, terminal operators should consider

investing in the equipment that will increase the handling speed, and consequently

shorten the stay of vessels at berth.

Univariate and multivariate analyses were conducted on the data obtained

from Part B of the questionnaire. Although the three private terminals KICT,

PICT and QICT offer the same kind of services, the result of one-way ANOVA

shows that there is nevertheless a significant difference in the opinion of respondents

about the performance of the four terminals, in terms of handling

rate, vessel stay, cargo loss and damage, equipment availability, large cargo

handling, asset specificity, personal contacts and container search. There is

need for improvement in these areas to allow terminal operators to cope better

with the growing competition.

To conclude, this paper, to my knowledge, is the first and the only analysis

of the selection criteria of international carriers when choosing container

terminals in Pakistan. It provides useful information for the port authorities and

terminal operators for the understanding and fulfilment of the requirements of

international carriers.


Journal of the Eastern Asia Society for Transportation Studies, Vol. 6, pp. 907 - 919, 2005


Hui-huang TAI &Cherng-chwan HWANG

This paper starts with an overall analysis of the performance of hub ports and change of trunk routes deployment based on statistical data of port operations, slots supply and operating routes of main shipping lines etc., collected from various sources. Secondly, the influential factors of selecting hub ports by shipping lines are investigated by conducting a questionnaire survey. Finally, the Gray decision model would be used to rank the relative competitiveness of these major ports from the viewpoints of container liners.

Since East Asia is the intermediate district connecting T/P and F/E routes, all the major shipping lines have expanded their service networks and size of fleet in this booming area, in order to increase their carrying capacity to meet the growing demand of container transportation.


In order to gain more insights about the relative importance of many possible influential factors on hub ports choice of major liners, a questionnaire survey was conducted, and the results are briefly presented in this section. Some of the data collected from this survey are also used to rank the relative of competitiveness among Shanghai, Kaohsiung, Yantian and Hongkong by the Gray Decision theory in next section.

3.1 Questionnaire Design and Survey

The questionnaire form is divided in two parts. Part 1 lists possible influential factors, in which score of each factor would be filled by the persons being surveyed. Factors are selected on the basis of literature review as well as discussions with experts in shipping industry, which are classified into following three categories.

Category A: internal factors of port, including efficiency of handling facilities, area of

marshalling yard, total no. of berth, draft of harbor, level of port charge, type of port authority, and quality of customer service. These 7 items are coded as A1-A7.

Category B: external factors of port, including cargo source of hinterland, efficiency of clearance, location of port, convenience of inland transportation, and frequency of trunk and feeder routes. These 5 items are coded as B1-B5.

Category C: operational factors of shipping lines, including saving in operating cost, preference of mother port, political considerations, capability of branch/agent, coordination of shipping alliance, and investment of dedicated terminals. These 6 items are coded as C1-C6.

The score of each factor is measured by Likert's 5 scale of rank, in which rank 5 to rank 1 represents "very important", "important", "average", "less important", and "not important" respectively.

Part 2 of the questionnaire form collects basic operational information of the shipping firms being surveyed, which are summarized in Table 7 and Table 8.

The questionnaire form is sent out to all the top 20 shipping lines, their branches and key

agents spreading in all of the 4 hub ports, including portage and shipping agents, but

excluding the freight forwarders that are not dedicated shipping agents of any liner. The

survey was conducted during the September, 2004. A total sample of 146 copies of the

questionnaire has been sent out, and 91 returned copies of which are examined as valid.


Based on the analysis of collected statistics, there are heaviest container flows between the East Asia and North America in the global liner shipping market. The major shipping lines expand their service network of trunk and feeder routes among several hub ports, put into large amount of fleet capacity in this region. It is clear that there exist competition among hub ports in this region, especially among Shanghai, Kaohsiung, Yantian and Hongkong.

The result of questionnaire survey indicates that among the 18 influential factors that might affect liners' hub ports choice, the most important factors are "handling efficiency" and "draft of harbor" that belong to the internal factors of ports, "cargo source of hinterland" and "frequency of routes" that belong to the external factors of ports, and the "saving in operating cost" for shipping lines. For the relative competitveness among these ports, Hongkong is the one with the highest competitiveness, followed by Kaohsiung. But more sophisticated method in company with more data collection work is recommended for further study, to obtain more persuasive result.


Port choice and freight forwarders

Jose L. Tongzon, a, ,

aGraduate School of Logistics, Inha University, 253 Yonghyun-Dong, Nam-ku, Incheon 402-751, South Korea

Received 8 July 2007; 

revised 17 February 2008; 

accepted 20 February 2008. 

Available online 29 April 2008.

4. Conclusion

This paper has sought to determine the key factors in port selection and assess their relative importance, using a survey method applied to a sample of freight forwarders. The findings suggest that such factors as high port efficiency, good geographical location, low port charges, adequate infrastructure, wide range of port services, connectivity to other ports, adequate infrastructure and others are important in the port selection process. Their relative importance, however, differ, with port efficiency considered as the most important factor. This finding is consistent with the recent study by Ugboma et al. (2006) in the context of Nigerian ports, which further reinforced the high importance shippers attach to port efficiency in their port choice decisions. In particular, these findings provide an empirical support that port efficiency is the most important factor in the port selection from the perspective of the freight forwarders. It is, therefore, essential that port operators and authorities give top priority to improving their overall level of efficiency relative to other factors in order to attract more freight forwarders to use their ports. In exploring the decision-making style and port selection process, the survey confirms the sequential decision making process resembling the findings made by (D'Este and Meyrick, 1992) and (D'Este, 1992) in their studies of shippers purchasing shipping services across the Bass Strait.

It further supports the hypothesis that the freight forwarders' port selection is not a simple but a complex and two-stage process which takes into consideration factors other than the conventional factors used in the traditional port choice models. To a certain extent the findings in this study therefore lend empirical support to the proposition that ports are not viewed by the freight forwarders in isolation but are considered together with other requirements associated with the movement of cargoes across the port-oriented supply chain. It therefore supports the new approach that models port choice within the framework of a port as an element of a supply chain which can provide us with a better understanding of the determinants of ports choice. Since most freight forwarders choose the shipping line first and then choose the port from those served by the shipping line, it is also important that port operators and authorities should pay special attention on how to attract shipping lines to call at their ports.

Although the survey has been limited to a sample of freight forwarders in Southeast Asia and selected ports, the results provide a useful empirical contribution to this increasingly important issue of port choice in this increasingly competitive trading environment in the context of a scant literature on port choice and offer an additional basis for further study into port choice and performance from the freight forwarders' perspective within the overall supply chain. In addition, this study complements the existing studies on the decision-making process of port users by examining the case of freight forwarders based in Southeast Asia which has not been undertaken before as previous studies have focused mainly on port users based in North America, Europe and Australia. Moreover, there is very little investigation done on the port selection process of freight forwarders and on the link between port choice determinants and performance (Ng, 2006).


Port Choice Determinants in a Competitive Environment*



Associate Professor

Department of Economics,

National University of Singapore

Kent Ridge Crescent, SINGAPORE 119260

Email: [email protected]

Phone: 65-874 6258

Fax: 65-775 2646

Revised: 13 September 2002

Based on the related literature survey, the following factors are postulated to have a significant impact on the choice of ports amongst shippers located at the industrial centers of Malaysia, Singapore and Thailand: frequency of ship visits, efficiency, adequacy of port infrastructure, location, competitive port charges, quick response to port users' needs and port's reputation for cargo damage.

Frequency of ship visits

greater frequency of ship calls allows for greater flexibility and lower transit time. Thus, the more ship visits a port has, the more attractive it is to shippers.

Port efficiency

Although frequency of ship calls is a significant factor for shippers in port choice, ports can also attract shippers due to their high levels of efficiency.

Adequate Infrastructure

Infrastructure in its widest context refers not simply to the number of container berths, cranes, tugs and terminal area, but also to the quality of cranes, quality and effectiveness of information systems, availability of inter-modal transport (such as roads and railways), the approach channel provided and the preparedness or otherwise of the port management

If the volumes handled far exceed a port's cargo-handling capacity, this will result in port congestion and inefficiency, and thus can turn off port users. Furthermore, limited access to current information about shipment arrivals due to lack of adequate information system will slow the documentation process and thus the smooth functioning of a port. Without adequate inter-modal links, shipper cannot easily move cargo to and from the port, which could lead to congestion, delays and higher costs.


Conventional notions of port choice have focused on geographical location as one of the main determinants of a port's attractiveness. The choice of a port is not merely a function of proximate convenience but derives considerable implications as well from the overall transit costs of cargo trafficking. For example, the distance between the port and the shipper's premises has a major impact on inland transportation costs.

Port charges

There are different types of port charges, which vary between ports in terms of levels and structures depending on the nature and functions of ports. Except for landlord ports, which derive their revenues from rents, port charges are generally levied on the basis of port visits and/or cargoes. Examples of ship-based types include port navigation fees, berthage, berth hire, harbour dues and tonnage while cargo-based types include wharfage and demurrage. Berth hire and berthage are usually levied either on the basis of net registered tonnes (NRT) or against gross registered tonnes (GRT). Stevedoring and terminal handling charges are levied on cargoes with different rates for different cargoes. Direct port charges may eventually be reflected in the freight rates shippers have to pay. Other types of costs which shippers eventually pay include ancillary charges such as costs of pilotage, towage, lines, mooring/unmooring, electricity, water and garbage disposal. Previous studies produced varied findings on the relative importance of port charges as a determinant of port choice.

Quick response to port users' needs

Ports are also expected to respond quickly to port users' needs. This means that ports would have to constantly monitor and understand the needs of port users in order to devise the quickest way to respond to them. Regular dialogues and social interactions between the port's public relations staff and the port users are quite useful in this regard.

Port's reputation for cargo damage

Perception of cargo safety can be more powerful and important than the actual safety. If a port has a reputation that the handling of cargoes is unsafe, this could drive away potential clients and discourage existing clients. Thus, marketing and promotional efforts by port authorities to highlight the port's positive characteristics and accomplishments could improve the port's reputation. A record of accomplishments and achievements gives assurance to customers in terms of quality and reliability. The latter is eminent for influencing carrier's choice of port as it is often the relative perception of customers that supersedes the actual port performance.

1.2 Relative importance of port choice factors identified

To further investigate the port decision factors and their relative importance to shippers, a survey was conducted among a sample of freight forwarders located at the industrial centres of Malaysia (Penang), Singapore and Thailand (Bangkok). Penang was chosen over Kuala Lumpur because the freight forwarders here have a more choice of which port to use, given the freight forwarders' closer proximity to the major ports of Malaysia and the port of Singapore. A sample of 47 major freight forwarders based in Penang of Malaysia, Bangkok of Thailand and Singapore were covered by a questionnaire survey. These freight forwarders were randomly chosen from the list of freight forwarders, provided by the Association of Freight Forwarders in Malaysia as well as from the business directories of Malaysia, Singapore and Thailand. Questionnaires were first mailed to the randomly selected freight forwarders before personal interviews were conducted to clarify certain points with those respondents who had unclear responses for certain items. The response rate was only about 25 percent.

The first set of statements refers to how port users view the importance of price in relation to the quality of service in the port choice process. The responses strongly favoured the quality of service over price, although there is a maximum price that they are willing to pay. Thus, as long as the price is below this limit, the quality of service takes precedence. About 76 percent of the respondents agreed that there is a minimum level of service on which they won't compromise. Only about 32 percent agreed that a low price can compensate for an inferior level of service. Further, almost 81 percent have agreed that a record of frequent delays in shipment would disqualify a port from consideration for future contracts.

Shippers are conservative decision makers. When given a choice between a conservative decision and a potentially profitable but risky one, they would take the conservative option. Sixty-six percent of shippers expressed a preference for the conservative one, while more than 80 percent of them would see no need to change ports, if the current port is performing satisfactorily. All of the shippers have agreed that preserving the reputation of their company and the goodwill of clients is the most important consideration in the choice of ports.

The shipper responses have also provided some insight into the mechanics of their port selection process. It is evident from their responses that the predominant approach is first to determine which ports can provide the required service and then to eliminate successively inferior options. Eighty-seven percent of shippers stated that they would only consider ports that provide the required service. Seventy six percent have agreed that they weigh up all the advantages and disadvantages of all the ports that might be capable of providing the service. In making a choice the bulk of the shippers (68.1 percent) have relied on personal contacts, knowledge and experience. In the sequencing of choices, 74.5 percent choose the shipping line first and then choose the port from those served by the shipping line. Only 23.4 percent decide the port to ship from before selecting the shipping line.

3 Conclusion

This paper has sought to determine the key factors in port choice and to assess their relative importance, using a survey method applied to a sample of shippers and basic econometrics. The findings of the survey suggest that such factors as high port efficiency, shipping frequency, adequate infrastructure, good location, low port charges, quick response to port users' needs and good reputation for cargo safety are the key determinants in the port selection process. Their relative importance, however, differ, with port efficiency considered as the most important factor. The decision-making style and port selection process also resembles the findings made by D'Este and Meyrick (1992) and D'Este (1992) in their studies of shippers purchasing shipping services across the Bass Strait. The prime importance of port efficiency, particularly in the cargo-handling aspect, is further reinforced by the econometric findings. Although the econometric estimation was based on a small sample, the results do provide some indication of the importance of port efficiency in port throughput determination.

These findings are important because in an increasingly competitive port environment, it is essential to know the key factors that come into the decision process of major port users and their mode of port selection. In particular, these findings provide an empirical support that port efficiency is the most important factor in port selection and throughput determination from the perspective of the shippers. It is, therefore, essential that port operators and policy makers give top priority to improving their overall level of efficiency relative to other factors in order to attract more shippers to use their ports. Hence, Malaysia's threat to the port of Singapore as the region's hub port can only become more real if the Malaysian ports can match Singapore's high level of port efficiency. Further, in exploring the concept of services provided and its evaluation from port users' perception, additional insights into port users' satisfaction and hence improvement measures are provided. It is their perceptions, their decisions and their behaviour that could matter.


Port competition and selection in contestable hinterlands; the case of Austria

Peter W. de Langen

EJTIR, 7, no. 1 (2007), pp. 1-14

Received: October 2005

Accepted: October 2006

This paper deals with port competition and port selection for cargo to/from Austria. Austria is located centrally in Europe and seaports in at least five countries are used for imports and exports.

In this paper an analysis is made of port competition and port selection in Austria, a country that is served by at least six different ports in five different countries: Hamburg and Bremen (Germany), Rotterdam (The Netherlands), Antwerp (Belgium), Trieste (Italy), and Koper (Slovenia). The analysis of port selection processes in a region with fierce competition from ports in different countries adds to existing studies of port selection.

6. Conclusions

In this paper an analysis was made of port selection and port competition for cargo to/from Austria. Austria was selected because this country is clearly a 'contestable hinterland' for various European ports. The analysis of Austria has yielded four relevant research findings.

First, the year to year changes of market shares of competing ports are considerable. This demonstrates that cargoes do shift between ports and that market share can be lost or won. This may be attributed at least partially to fierce port competition. Thus, ports constantly run the risk of losing market share in their contestable hinterland. More research in this direction is required, and perhaps more detailed data like market shares for specific commodities could be collected to expand this study. A further research question with regard to switching of port users between ports is 'can ports differentiate prices between contestable and captive customers?'

Second, the opening of the Rhine-Main-Donau Canal considerably improved the competitive position of the ports of Rotterdam and Antwerp, especially for low value commodities. The increase of market share of both ports did not take place in one or two years, but took a whole decade. Since the survey results demonstrate shippers and forwarders make ´rational´ port choices instead of choices based on tradition or personal contacts, this long period of ´adaptation´ to new transport options is explained by the existence of substantial 'friction costs' (see Hesse and Rodrigue, 2004). These friction costs prevent instantaneous shifts or cargo flows to another inland mode and port. This is a relevant research finding for other ports that invest in new infrastructure and leads to the interesting question what actors in ports can do (if anything) to accelerate the shift of cargo to their port. One important factor in this respect may be the problem that scale economies that can lead to lower transport costs do not arise spontaneously, but require 'collective action' (see De Langen and Chouly, 2004).

Third, the demand for port services of forwarders seems to be more price elastic than the demand of shippers. Due to the low response rate this conclusion cannot be drawn with certainty, but this tentative conclusion provides a basis for further research. Since forwarders control a large share of transport flows (see Murphy and Daley, 2001), research on the behaviour and port choice decisions of forwarders is relevant. The effects of the terminal handling charges on port choice deserve particular attention16 (see Fung et al., 2003).


Maritime Economics & Logistics (2009) 11, 260-269. doi:10.1057/mel.2009.9

Port selection from a hinterland perspective

Lorena Garcia-Alonsoa and Joaquín Sanchez-Sorianob


This article deals with the analysis of port selection from a hinterland perspective. Attending to the suggestions of D'Este (1992) and Mangan et al (2001), the study of port selection is made in a holistic way, focusing on the actual inter-port container traffic distribution. The results show that the port-province distance remains a relevant variable in the port selection process (even for container traffic) despite all the transport sector improvements, and it confirms the conclusion of Sargent (1938): cargo tends to seek the shortest route to access the sea. Hence, the concept of hinterland does contribute to explain the evolution of the activity of a port, in spite of the development of the inter-modal transport and the increase in inter-port competition. Our hypothesis is that the hinterland distance is still a variable so important that firms, when deciding about their location, take into account the location of the ports offering the services they need, whereas firms already established tend to choose the services offered by the nearest port.

Nevertheless, it is also important to keep in mind that the contribution of their own hinterlands is not enough to explain the success of the ports of Algeciras, Barcelona and Valencia during the last decade. According to Gouvernal et al (2005), the development of new logistics chains using the Mediterranean Sea has been very important for the success of these ports. On the other hand, the growth of the activity of the port of Bilbao has been considerably smaller. This port is located on the Atlantic coast, where the evolution of container traffic (and the development of supply chains) did not favour its activity. Consequently, the evolution of the traffic of the port of Bilbao is much more linked to its own hinterland than in the other ports.

It seems, therefore, that the evolution of the port activity matches the strategic position of the ports according to the terms introduced by Fleming and Hayuth (1994): centrality and intermediacy. That is, it depends on the dynamism of the hinterland of the ports and on their inclusion in the routes of shipping lines. We can conclude that both perspectives, the maritime (usual in the literature) and the hinterland (proposed in this article), complement each other. Then, both perspectives are necessary to analyse the evolution of the activity of medium ports: ports included in important maritime lines, but with an important volume of national traffic.

Top of page


1 We observe that the value of the Cramer coefficient V for the model with only the distance is better than the corresponding for the better model from the point of view of the likelihood. The reason for this is the Maximum Likelihood Estimators (MLE) are efficient and asymptotically unbiased but not robust, so they can be strongly influenced by outlier observations, as it is the case in this article.


Maritime Economics & Logistics, 2003, 5, (23-39)

Shippers' Port and Carrier Selection Behaviour in China: A Discrete Choice Analysis

Piyush Tiwari, Hidekazu Itoh and Masayuki Doi


The analysis of shippers' behaviour with respect to choice of ports and carriers is essential for policy formulation regarding improvement and development of port infrastructure. This is one of the few studies on the subject that attempts to model this behaviour by using an empirical model, and probably the first attempt to model the joint choice of carrier and port in China. The data used are unique and come from a survey of shippers conducted by The International Centre for the Study of East Asian Development, Japan, for the year 1998. Earlier research on the choice of carrier indicates that service factors and costs are important parameters in determining choice. This paper moved a step further and tried to estimate how the market share of various port-carrier combinations would change in response to changes in their key variables.

The results indicate that Chinese shippers and forwarders are conservative and prefer Chinese shipping lines primarily because they have larger fleets catering to China and longer relations established over a long period of time. Shippers are indifferent to foreign shipping lines and their choice is driven mostly by the port they would like to use to import or export cargo. However, foreign shipping line operations are relatively recent in China and a change in shipper behaviour towards their use can be expected in the near future. The number of TEUs handled in a port indicates congestion, and has a negative impact on shippers' decisions. The number of berths, and fleet size enhance efficiency in moving cargo and have positive coefficients.

The port distance from a shipper's location is an important variable determining port choice. Distance has negative elasticity.

The estimated model is used to determine market share elasticities. These are important policy parameters explaining changes in the market shares of various alternatives in response to 1% change in a policy variable. For example, they estimate what would be the impact on the market share of a port, vis-à-vis other ports, if the shipper's distance from this port increases by 1%. Results are quite illustrative. An increase in the distance of a shipper from Dalian by 1%, assuming the shipper uses a Chinese carrier, reduces the market share of this combination by 7.9%, while the market share of all other port-carrier combinations increases by 0.95% each. If the shipper uses a Southeast Asian carrier instead, the market share of this 'Dalian-Southeast Asian carrier' combination decreases by 8.62%, while the market share of all other port-carrier combinations increases by 0.23%.

The 'fleet size' elasticity indicates that shippers are sensitive to changes in the number of vessels of Chinese shipping lines. An increase in the number of vessels of Chinese shipping lines by 1% increases the market shares of those alternatives by around 5.4%-6.1% depending on the port used.

Top of page